API Documentation

Get up and running in under a minute.

Every API request requires one of the following authentication methods.

Base URL

All vision endpoints accept image uploads and return structured JSON results. Click each endpoint for parameters, code examples, and response schemas.

Process video files and live RTSP streams in real-time. Available on Pro and Enterprise plans.

# Start stream via API
GET https://api.vyrion.ai/v1/video/stream?source=rtsp://camera:554/stream&task=detect

# Or use the Playground → select RTSP → enter URL → press play

curl -X POST https://api.vyrion.ai/v1/video/upload \
  -H "X-API-Key: vyr_live_YOUR_KEY" \
  -F "file=@footage.mp4" \
  -F "task=detect"

Create no-code computer vision pipelines by chaining multiple AI models together. Combine detection, segmentation, classification, OCR, and enhancement in a single workflow.

Pre-configured AI solutions for specific industries and use cases. Each playground is optimized with domain-specific models.

AI-powered medical imaging analysis for research and diagnostic assistance. Supports DICOM and standard image formats.

Generate professional PDF and HTML reports from any vision analysis result. Includes detection summaries, images, statistics, and shareable links.

Discover community-published models or share your own custom-trained models and earn credits.

Use our API directly with your preferred language.

Python

import requests

API_KEY = "vyr_live_YOUR_KEY"
BASE = "https://api.vyrion.ai"

# Object Detection
def detect_objects(image_path: str, confidence: float = 0.25):
    r = requests.post(
        f"{BASE}/v1/detect",
        headers={"X-API-Key": API_KEY},
        files={"file": open(image_path, "rb")},
        data={"confidence": confidence, "prompts": "person,car"}
    )
    return r.json()

result = detect_objects("scene.jpg")
for d in result["detections"]:
    print(f"{d['class_name']}: {d['confidence']:.2%}")

JavaScript

const API_KEY = "vyr_live_YOUR_KEY";
const BASE = "https://api.vyrion.ai";

async function detectObjects(file: File) {
  const form = new FormData();
  form.append("file", file);
  form.append("confidence", "0.4");

  const r = await fetch(`${BASE}/v1/detect`, {
    method: "POST",
    headers: { "X-API-Key": API_KEY },
    body: form,
  });
  return r.json();
}

All plans include access to the same powerful AI models. Credits reset monthly.